In an experiment seemingly ripped from Law and Order or CSI, UK Psychologists Rob Jenkins and Christine Kerr demonstrated that it is possible to identify the faces of unseen bystanders in a photograph via the reflections in the eyes of the photographed individuals. Jenkins and Kerr took portrait pictures of individuals while other participants stood out of the photograph but in the line of site of the photographed subject. The researchers then used photo-editing techniques to extract faces in the reflections of the eyes of those photographed. Using these extracted facial images, they conducted two tests of identification: a face-matching task and a spontaneous recognition task. In both tasks participants were able to identify those in the picture at rate better than chance. Given the well-documented human ability to recognize faces, even in poor quality images, the result is not especially surprising. What is striking, however, is the original source of the image, and the fact that despite being 30,000 times smaller than the subject’s face in the original photograph, the extracted faces were still identifiable–which is a testament to both modern photographic technology and the human ability to recognize faces. Generalizing their work, Jenkins and Kerr highlight the applicability of this technique to aid in criminal investigations, where film and camera are recovered as evidence.
- Jenkins R, Kerr C (2013) Identifiable Images of Bystanders Extracted from Corneal Reflections. PLoS ONE 8(12): e83325. doi:10.1371/ journal.pone.0083325 (Full Study)